Body Dysmorphic Disorder: Differences in Age and Compulsive Online Behavior in a Swedish Sample
Sofia Ersson & Rebecca Holvik
Örebro Universitet
Abstract
Compulsive repetitive behaviors and mental acts due to concerns about your appearance, are symptoms of body dysmorphic disorder (BDD). Previous research suggests that the compulsive behaviors found in people with BDD occur in online and offline settings (e.g., extensive editing of selfies intended for publication online and excessive mirror gazing offline). Also, previous research shows that BDD and social media use vary with age. Therefore, the current study aimed to examine age as a moderator in the relationship between compulsive behaviors online and the risk behaviors of BDD offline, through a cross-sectional design. The inclusion criteria for the study were being a minimum of 16 years old, a Swedish citizen and a user of social media. The data were collected through a survey, consisting of questions intended to screen for compulsive behaviors online in relation to appearance concerns, risk behaviors of BDD offline and the prevalence of BDD. The results showed that younger participants engaged in more compulsive behaviors online and risk behaviors of BDD. The group in high risk of BDD also engaged more in both behaviors, than participants in low risk of BDD. In addition, the results showed that age did act as a moderator in the relationship between compulsive behaviors online and risk behaviors of BDD. Age showed to especially affect the relationship between the number of compulsive behaviors online and risk behaviors of BDD in older participants.
Key words: Body dysmorphic disorder (BDD), Social media use, Compulsive behavior, Age
Dysmorfofobi: Skillnader i ålder och tvångsmässigt beteende online i ett svenskt urval Sofia Ersson & Rebecca Holvik
Örebro Universitet
Sammanfattning
Tvångsmässiga repetitiva beteenden och mentala handlingar som beror på oro kring sitt utseende, är symptom av dysmorfobi (BDD). Tidigare forskning visar att tvångsmässiga beteenden sker både online och offline (tex., överdriven redigering av selfies i syfte att publicera dem online och att spegla sig överdrivet mycket offline). Tidigare forskning visar även att BDD och sociala medier-användning varierar med ålder. Genom en tvärsnittsdesign ämnade därför denna studie att undersöka om ålder modererar relationen mellan tvångsmässigt beteende online och riskbeteende för BDD. Datainsamlingen bestod av ett frågeformulär med frågor kring tvångsmässigt beteende online i relation till oro kring utseendet, riskbeteende för BDD samt prevalensen av BDD. Inklusionskriterierna för att delta i studien var att vara minst 16 år gammal, svensk medborgare samt användare av sociala medier. Resultaten visade att de yngre deltagarna i studien utför både fler tvångsmässiga beteenden online samt fler riskbeteenden offline än de äldre deltagarna. Den grupp som ansågs ha hög risk för BDD utförde även båda beteendena mer än de deltagarna med låg risk för BDD. Resultaten visade vidare att ålder var en moderator i relationen mellan tvångsmässiga beteenden online och riskbeteenden för BDD. Mer specifikt hade en äldre ålder störst effekt på relationen mellan beteendena online och offline.
Nyckelord: Body dysmorphic disorder (BDD), Sociala medier-användning, Tvångsmässiga beteenden, Ålder
Handledare: John Barnes Psykologi kandidatkurs
Body Dysmorphic Disorder: Differences in Age and Compulsive Online Behavior in a Swedish Sample
An unwarranted and excessive amount of concern about your appearance can have a negative impact on your life and you may be at risk of having body dysmorphic disorder (BDD). BDD affects an individuals’ life in a debilitating way because of obsessive thoughts, preoccupation and compulsive behaviors that can take up to 8 hours per day (American Psychiatric Association [APA], 2013). The compulsive behavior has been found to take place in both offline and in online settings (Petersson & Tuupanen, 2020). Subsequently, it was also found in the same study that these compulsive behaviors online may predict a risk of BDD. Meanwhile, statistics show that we are more online today than previously, and the younger age groups are the most common users of social media in Sweden (Statistiska Centralbyrån [SCB], 2020). Research has also suggested that BDD may be more common in younger age groups than in older age groups (Möllman et al., 2017). Hence, the present study explored if more compulsive behaviors online were related to an increase of risk behaviors of body dysmorphic disorder, and if that relationship was affected by age.
This paper will give the reader an understanding of the psychological diagnosis of body dysmorphic disorder by first introducing a summary and previous research on the disorder. Secondly, a presentation of the current knowledge about social media use and compulsive behaviors online will follow. Third, it will be described how online compulsive behaviors, risk behavior of BDD, risk of BDD and age were measured in the current study. Lastly, the authors will present the results of the current study, followed by a discussion about the method and the results found in the present study.
The Psychological Diagnosis of Body Dysmorphic Disorder
Body dysmorphic disorder revolves around a self-perceived defect or defects in one's appearance. BDD is a psychiatric disorder that involves a preoccupation with one or more
perceived defects in one’s appearance, that are usually invisible to others. Although, if the defect is visible, then it’s usually very minor (APA, 2013). The areas of preoccupation in one’s appearance in BDD, can also change over time (Brohede et al., 2016). In addition, the number of areas that people with BDD have experienced as defected can include up to seven different areas (Phillips et al., 2006). Hence, BDD can involve preoccupation about more than one perceived defect that may not be visible to others at all.
Different areas of the body can be perceived as defected by people with BDD and the affected areas can also differ between genders. It has been shown to be common for males to have concerns regarding the appearance of their genitals and, they have also shown to be more prone to be concerned about their body build being either too small or not muscular enough (Phillips et al., 2006). The preoccupation with concerns about body build is a form of BDD known as muscle dysmorphia and it can be combined with other appearance concerns (APA, 2013). In contrast, according to APA, women with BDD are more common to have appearance concerns regarding their skin, breast, stomach, weight, thighs, buttocks, toes and legs. Accordingly, the bodily area of concern to people with BDD can differ depending on gender.
Advances in the understanding of body dysmorphic disorder have caused changes regarding the classification of the disorder. The Diagnostic and Statistical Manual of Mental Disorders (DSM) is a manual for psychiatric disorders including body dysmorphic disorder (APA, 2013). In the fifth version of the manual, BDD is found in the obsessive-compulsive and related disorders' chapter, from previously being characterized as a somatoform disorder. APA (2013) characterizes all disorders that share the same symptomatic aspect of obsessive-compulsive tendencies, as an obsessive-obsessive-compulsive and related disorder. In BDD, the obsessive-compulsive tendencies are described as being limited to preoccupations regarding
appearance (APA, 2013). Hence, BDD is a psychiatric disorder including obsessive-compulsive tendencies in connection to their appearance concerns.
The obsessive-compulsive tendencies found in people with BDD have been included in the diagnostic criteria of the disorder. The diagnostic criteria of BDD in DSM-5 includes a criterion that refers to the compulsive tendencies of BDD as repetitive aspects in terms of behaviors and mental acts (APA, 2013). Also, it’s specified that this behavior needs to take place at some point during the course of the disorder to reach the full criteria of BDD. APA (2013) describes these behaviors and mental acts as intrusive, very hard to resist and time-consuming. Thus, compulsions regarding behaviors and mental acts in connection to appearance concerns are part of the symptomatic presentation of BDD.
The compulsive repetitive behaviors and mental acts that people with BDD engage in takes up a lot of time and can be distressing. According to APA (2013), the repetitive
behaviors and cognitions found in people with BDD can take up approximately 3-8 hours each day. Repetitive behaviors and mental acts that are common for people with BDD can include repeatedly checking in mirrors, picking skin, comparing yourself to others and seeking reassurance from others in regard to the appearance concerns (Lambrou et al., 2012). Moreover, safety behaviors have also shown to be common in BDD where the person may camouflage their perceived defects with certain clothes or by using make-up. People with BDD have also shown to engage in excessive grooming, e.g., combing, styling or shaving hair, among other things (APA, 2013). In addition, the time people with BDD spend on engaging in these behaviors and mental acts can cause impairment and distress. Thus, people with body dysmorphic disorder repeatedly and compulsively engage in distressing checking behaviors because of their appearance concerns.
How much awareness someone with BDD has in regard to knowing if their perceived defect is real or not can vary. Since the defect in people with BDD usually is invisible to
others, a specifier for insight has been included in DSM-5 for BDD (APA, 2013). The insight specifier ranges from grading someone to have a “good/fair insight” meaning that the person knows that their perceived defect or defects probably aren’t as bad as they believe it to be, to “absent insight/delusional beliefs” where the person is convinced that their perception of their defect or defects are completely true (APA, 2013). Mostly, insight has shown to be poor in people with BDD (Phillips et al., 2012). Hence, delusions regarding the perceived defect in someone with BDD are common, where they may truly believe that they are defected or deformed in some way, even if the defect is invisible to others.
Body dysmorphic disorder can co-occur with other psychiatric disorders. Similarities and comorbidity are common between obsessive-compulsive disorder (OCD) and BDD (APA, 2013; Sharma et al., 2019). Where, similarities are for example that both disorders are known to present clinical symptoms of compulsive intrusive thoughts and delusional aspects. Although, people with BDD have shown higher degrees of delusionality and suicidal
tendencies than people with OCD (Phillips, et al., 2012; Sharma et al., 2019). Apart from this, BDD have also shown to present comorbidity with major depressive disorder and social phobia (Veale et al., 2016). Additionally, concerns regarding weight have been described as mainly being a concern in eating disorders, but it may also occur in individuals with BDD (APA, 2013). Further, both eating disorders and BDD can occur at the same time. Hence, people with BDD may present additional clinical symptoms than what is caused by BDD, due to another comorbid disorder.
People with BDD rarely seek professional mental help for their appearance concerns. One study found that it was almost as common to seek information about cosmetic surgery as it was to seek professional mental help among participants with BDD (Schulte et al., 2020). Also, more participants used the internet to seek information about their appearance concern rather than seeking treatment by clinicians. Instead of seeking professional mental help for
concerns related to BDD, people with BDD may try to improve their imagined flaws in different ways, e.g., plastic surgery (APA, 2013; Brohede et al., 2016; Phillips, 2005). This is also supported by another study, which found that it was more common among people with high risk of BDD to seek treatment through plastic surgery than by seeking psychological care for their appearance concerns (Petersson & Tuupanen, 2020). Thus, people with BDD or people in high risk of BDD often seek help by other means than by psychological care for their appearance concerns.
There has been shown to be several reasons for why people with BDD don’t seek psychological care. Common causes for why people with BDD don’t seek help by
psychologists have been reported to be due to dissatisfaction with previous treatment; the perception that the individual did not need any psychological treatment; that cosmetic or medical treatments were preferred; or that the individual did not know who could help with his or her concerns (Brohede et al., 2016). However, the same study found that the most common causes are due to stigma and shame, where the individual reported feeling ashamed about his or her concern or concerns regarding their appearance. Phillips (2005) describes BDD as a “secret disorder” because a lot of individuals never express their concerns to health care professionals, family or friends. Among some people, the shame they felt about their concern decreased when they had been diagnosed with BDD (Brohede et al., 2016). The same people also reported that they felt an ease because they knew that they could get treated. In conclusion, feelings of embarrassment are one of many reasons for why people with BDD seek help by other means than psychological care but, if they do seek psychological help the feelings of embarrassment related to the disorder may subside.
BDD is a rather unknown disorder. Actually, very few clinicians are aware of BDD (Brohede et al., 2016). It has shown to be common that people with BDD aren’t diagnosed with the disorder, although appearance concerns are their main problem (Phillips, 2005).
Further, the patients have sometimes been diagnosed with a co-occurring diagnosis instead of BDD. In some cases, the patients were more aware of the disorder than the clinicians,
resulting in the patient having to inform the professional about the disorder (Brohede et al., 2016). Thus, the unawareness about BDD contributes to underdiagnosis of the disorder.
Implications from the severity of symptoms of body dysmorphic disorder can lead to a life of poor quality. Impairments to work and academia are common consequences of the appearance concerns in BDD (Phillips et al., 2006). In addition, social avoidance has also been shown to be a common consequence due to symptoms of BDD, which can lead to the person becoming completely housebound. It has also been found that people who screened positive for BDD were more likely to be unemployed or on sick leave than the people who screened negative for BDD (Brohede et al., 2017). According to APA (2013), almost all people with BDD suffer from an impaired psychosocial functioning due to their appearance concerns. In addition, BDD has also shown to have a high risk of becoming chronic, if the onset takes place before age 18, although, the mean age of onset for BDD is around 16 years old and the most common age of onset is at 12 years old (APA, 2013). Furthermore, the distress and impairment in people with BDD caused by symptoms of the disorder can also lead to suicide attempts. Suicidal thoughts have also been commonly reported by people with BDD (Brohede et al., 2016; Möllmann et al., 2017). Thus, body dysmorphic disorder is a debilitating psychiatric disorder which can have impairing consequences to a person's life.
Previous studies on the prevalence of body dysmorphic disorder have shown that the disorder is relatively common in different age groups. A systematic review presented similar prevalence rates for adults (1.9%) and adolescents (2.2%) and, a somewhat higher rate for students (3.3%) in community samples (Vale et al., 2016). Similarly, to the prevalence found in students in the previous research, a recent study on adolescents and young adults including the ages of 15-21, also showed an elevated prevalence rate of BDD (3.6%) for this group
(Möllman et al., 2017). Also, the same study indicated that rates of BDD may be potentially higher in adolescent samples than what previous studies have found. Furthermore, one recent study found results regarding prevalence in an adult sample that are not in line with previous research. The study found a somewhat higher prevalence rate for adult females (3.4%) than what has been previously found for this age group in the general population (Petersson & Tuupanen, 2020). Although, among males, the same study found a prevalence of 0.8%. Apart from this, BDD also occurs among the elderly, although little is known about the disorder in this age group (APA, 2013). In conclusion, the prevalence of body dysmorphic disorder is quite common across different ages and, some studies indicate a higher prevalence rate of the disorder among adolescents and students.
Studies on prevalence of BDD in Sweden are limited. Although, some studies that have explored the prevalence of BDD in Swedish samples have shown that the rates may decline as age increases. The authors found three studies that have reported prevalence rates of BDD for Swedish community samples. One study on only females aged 18-60, reported a total prevalence rate of 2.1% (Brohede et al., 2015). When the same study divided the sample into three age groups, the prevalence of BDD showed to be highest in the youngest age group and lower in the older age groups. Similarly, another study also found the same pattern
regarding prevalence and age, with a sample consisting of both males and females aged 18-85 (Petersson & Tuupanen, 2020). More specifically, the prevalence of BDD was found to be highest in the youngest age group in that study as well. Apart from this, a longitudinal twin-study on adolescents and young adults in Sweden showed that BDD is quite common among the ages between 15-28 (Enander et al., 2018). The study reported a prevalence rate between 1 and 2%, where the prevalence was higher for the two cohorts with individuals aged 15 and 18, compared to the cohort with individuals aged 20-28. Concluding, since studies on
prevalence of BDD in Sweden are scarce, future studies are needed to further evaluate potential age difference in prevalence rates of BDD.
Social Media Use and Compulsive Behavior Online
The compulsive behaviors included in body dysmorphic disorder can also take place in an online environment. People with high risk of BDD have shown to engage in more compulsive behaviors online and also spend more time doing so, than people with low risk of BDD (Peterson & Tuupanen, 2020). Specifically, the same study also found that people with high risk of BDD engaged in more repetitive behavior such as comparing oneself to others online, more than people with low risk of BDD. Furthermore, a case report revealed that a person with BDD had engaged in problematic and excessive internet and Facebook use (Khanna & Sharma, 2017). Moreover, the person had also engaged in repeatedly taking an excessive number of selfies to manage her BDD-related distress. In conclusion, risk behavior of BDD such as compulsive behaviors do appear in online environments as well.
Social media use increases every year among Swedes and the younger population are the most frequent users. Social media use has grown with almost 20% in Sweden between 2011 and 2019, from 54% to 72% (Eurostat, 2020). During the years of 2018-2020, social media use was most common in the youngest age group (16-24 years) in Sweden (SCB, 2020). Furthermore, people aged 25-34 were the second most common users of social media. The use of social media appears to decline with age. Regarding the general internet use in Sweden during the year of 2020, the difference between the age groups is more subtle, with an overall everyday internet use of 88% of the population (SCB, 2020). Thus, social media use has grown rapidly among the Swedish population and has shown to be particularly popular by people aged between 16-34 years old.
Focusing on appearance and appearance concerns on social media relates to body image dissatisfaction. Among young women, photo-based activities are related to body
surveillance (Cohen et al., 2017). Body surveillance means to behave and think in a monitoring way, due to worry about one's appearance in the eyes of others. Among men, talking about appearance on social media was related to body surveillance, and the combined effect of talking about appearance and body surveillance led to body shame (Wang et al., 2019). Also, appearance comparisons online with another person was related to body image concerns (Seekis et al., 2020). A similarity between BDD and body image dissatisfaction is the repetitive aspect of comparisons to others (Ryding & Kuss, 2020). Thus, appearance-related behaviors on social media are connected to risk behaviors of BDD since they involve compulsive aspects in connection to appearance.
The negative impact of comparative behavior online may be explained by social comparison theory. The theory of social comparison infers that we engage in social
comparison to evaluate our opinions and abilities (Festinger, 1954). Recent research building on the theory of social comparison have also shown that engaging in upward social
comparison on Instagram had a negative impact on body image and resulting in more body dissatisfaction (Tiggeman & Anderberg, 2019). Also, the same study found that participants were more prone to engage in social comparison with others that were similar to themselves. Previous research has also found that appearance comparison with others was related to body dissatisfaction (Girard et al., 2018; Vartanian & Dey, 2013). As previously mentioned, it was also shown that appearance comparison online was related to body image concerns (Seekis et al., 2020). Hence, social media such as Instagram makes it possible to engage in comparative behavior that can impact our body image negatively.
There is uncertainty regarding what ages are included in certain age groups in research on BDD. One study conducted research on BDD on participants referred to adolescents and young adults and included participants ranging from 15 to 28 years old (Enander et al., 2018). Whereas another study included a sample consisting of 15 to 21 year old’s, also referred to as
adolescents and young adults by the authors (Möllmann et al., 2017). To contrary, one study which referred to their sample as adults consisted of individuals aged from 18 to 85
(Petersson & Tuupanen, 2020). In statistical reports on the use of social media sites among the Swedish population, the youngest age group ranged from 16 up to 24 and the following age group ranged from 25 to 34 (SCB, 2020). Therefore, due to different thoughts regarding what ages should be included in what age group in research on BDD, one approach could be to look at how age groups differ in their actual internet use.
Although findings have shown relations between body image concerns and social media use, there is little research on relations between BDD, internet use and social media use. The studies screening for BDD or risk behaviors of BDD and social media use known for the authors of the current study, was a case study and a master thesis (Khanna & Sharma, 2017; Petersson & Tuupanen, 2020). Both studies showed that people with BDD or people in high risk of having BDD, engage in appearance-focused behavior online, including
compulsive behaviors. Apart from this, a systematic review discussed similarities between BDD and body image concerns (Ryding & Kuss, 2020). The review suggests that risk behavior of BDD can be maintained through frequent social media use. Therefore, together with the fact that more time is spent on the internet, further studies about BDD and social media use are needed.
Aims of the current study
Regarding that the rates of BDD and social media use appear to vary with age, the research question of the current study was if compulsive behavior online is associated with risk behavior of BDD and if the relationship is affected by age. The current study aspired to contribute with more knowledge to the research field on compulsive behaviors online, such as on social media, in relation to body dysmorphic disorder in a Swedish context.Compulsive behavior online, risk behavior of BDD, and age were measured by an online self-report
survey. Also, the collected data for the current study is also part of a bigger research project on online behavior, eating disorders, body dysmorphic disorder and mental health that is out of scope for present study.
To examine the research question we had to first, measure if compulsive behavior online and risk behavior of BDD was associated with age. Next, we had to explore if the investigated behaviors also differed between age groups. Thus, the previously mentioned analyses made it possible to explore the research question of the current study which was to examine if compulsive behaviors online were associated with risk behavior of BDD, and if the relationship is moderated by age. In light of this, the hypothesis for the current study was that people who engaged in more compulsive behaviors online also would engage in more risk behavior of BDD and that this relationship would be affected by age.
Method Procedure
The current study was conducted in collaboration with a bigger international study on internet use and mental health. Together with the bigger research project, the authors of the current study constructed a web-survey which included eleven existing scales, aimed to collect data about online behavior and mental health. The scales included in the survey
screened for mental health issues such as depression, anxiety, obsessive-compulsive behavior, body image and different types of online behavior. Although, in light of the aims of the current study, only the data collected from the following three scales were used: The Body Dysmorphic Disorder Questionnaire (BDDQ) (Phillips, 2005), Appearance-Related
Repetitive Behaviors Online (ARBO) (Petersson & Tuupanen, 2020), and Body Dysmorphic Disorder-Symptom Scale (BDD-SS) (Wilhelm et al., 2016).
The current study used a cross-sectional method by collecting data through the online survey, which was constructed in a program called Artologik. To ensure the quality of the
survey, a small pilot sample consisting of six individuals was used to evaluate it before it was published. The pilot sample was recruited by convenience sampling. They were asked to report spelling mistakes and if they found any parts of the survey to be difficult to understand. The feedback from the pilot sample concluded in some minor corrections regarding the construction of the survey, but no major mistakes or difficulties were identified. Therefore, after some changes were made to the survey it was deemed acceptable to launch and start the data collection after being reviewed by the pilot sample.
The scales in the study were presented in threedifferent orders to be able to evaluate the presence of possible order effects that might affect the answers in the survey. Besides from the first presentation of the scales, the eleven scales included in the survey were then rotated two times during the period of data collection for the current study. The first rotation was done after approximately a hundred participants had submitted the survey and the next rotation was made after approximately another hundred participants had submitted the survey. Concluding, in three different orders of presentations of the scales among the participants of the current study.
The items within the scales were not randomized because they are included in validated scales and had to be presented in certain orders. Further, the questionnaire was written in English, because of the fact that the survey was also being part of a bigger
international research project which aimed to reach other nationalities than only Swedish.The period of the data collection for the currents study was two weeks between 11.24.2020 and 12.08.2020.
Participants
Snowball recruitment and convenience sampling were used to recruit participants for the current study. Inclusion criteria for the study were that participants had to be at least 16 years old, a Swedish citizen and a user of social media. The survey and an advertisement for
participation in the study was posted on Facebook and on Instagram. In addition, the ad to the survey was also posted in Facebook groups concerning people with BDD and in groups with other mental health concerns. The participants were also encouraged to distribute the survey link to others who could have been interested in participating, through the advertisement. Moreover, the link to the survey was also distributed to senior high schools, municipalities and companies within Sweden with the aim to reach a wider range of people in different age
groups.
The web-survey was completed by 243 people. Although, the final sample consisted of 206 participants due to some individuals being excluded since they did not meet certain inclusion criteria: 25 people were not Swedish citizens; one person was younger than 16 years old; and three people stated that they do not use social media. Also, eight participants were excluded because of a lot of missing data. Resulting, in a sample of 206 participants. One percent of the sample (n = 2) identified their gender as “other”, 26% as male (n = 54), and 73% as female (n = 150). The ages of participants ranged from 16 to 68 (M = 26.9, SD = 11). Table 1 further shows the sociodemographic characteristics of the current sample.
Table 1 Sociodemographic Characteristics of participants (n = 206). n % Gender Male 54 26% Female 150 73% Other 2 1% Age 16 - 24 110 53% 25 - 34 67 33% 35+ 29 14% Civil status Married 15 7% Single 106 52% Cohabitant 40 19% In a relationship 43 21% Other 1 1% Work status Student 131 64%
Full time employment 59 29% Part time employment 7 3%
On sick leave 4 2%
Unemployed 2 1%
Other 3 2%
Sick leave due to appearance concerns
Yes, I am on sick leave 7 3% Yes, previously 32 16%
No, never 165 80%
Treatment due to appearance concerns
Yes, surgical cosmetic 18 9% Yes, psychological 24 12%
Yes, medical 38 18%
The study sample was divided into age groups for some of the analyses to be able to explore potential age differences on compulsive behaviors online and on risk behavior of BDD. Due to the uncertainty concerning the definition of age groups in the literature of BDD, the authors decided to split the current sample into three groups based on the suggested statistics of social media use: the ages 16-24 in the first age group, 25-34 in the second age group, and participants 35 years or older in the third age group.
Measures
Compulsive behavior online, risk behavior of BDD and age were measured by a web-survey in the current study. The web-survey began with demographic questions about gender, age, geographical location, work status and civil status. There was also a question regarding whether the participant had been on sick leave due to appearance concerns and a question asking about if the participant had ever sought treatment or care due to appearance concerns. The scales included in the survey were: BDDQ, to measure high and low risk of BDD (Phillips, 2005); ARBO, which was used to measure the compulsive behaviors of BDD in an online setting (Petersson & Tuupanen, 2020); and BDD-SS, to screen for compulsive
behaviors that indicate risk behavior of BDD in an offline setting and (Wilhelm et al., 2016). Thus, the current study used the collected data from the BDDQ, BDD-SS and ARBO to measure the variables age, compulsive behaviors online and risk behavior of BDD. Body Dysmorphic Disorder Questionnaire
The Body Dysmorphic Disorder Questionnaire (BDDQ) was used to measure the estimated prevalence risk of BDD. BDDQ is a screening tool for BDD which is based on the DSM-IV criterions for the disorder (Phillips, 2005). Meaning, that the BDDQ can indicate the possible presence of BDD, but the results does not directly imply a diagnosis of the disorder. In addition, for a diagnosis of BDD to be confirmed, an interview by a clinician is needed to be able to confirm an individual's result on the BDDQ (Phillips, 2005). Furthermore, when
comparing the BDDQ judgement of BDD with clinician’s judgement of the disorder, the BDDQ will detect BDD correctly in 100% of the cases. The scale will also assess that there is no BDD similarly to clinician’s judgement in 89% of the cases. Thus, BDDQ screens for the risk of BDD and was therefore included in the web-survey.
The BDDQ is a self-report questionnaire consisting of eight questions, which is divided into four parts consisting of main questions and sub-questions. Although, there are four questions in the BDDQ that mainly distinguish if you are in risk of having BDD or not. According to Phillips (2005), an individual should be thought of as being at risk of having BDD if the individual answers yes to the question “Are you worried about how you look?” and, answers yes to at least one out of four sub-questions that ask if any certain area has been negatively affected by the preoccupation with their appearance, e.g., “Has it caused you any problems with school or work?”. In addition, it’s described that the preoccupation needs to take up at least 1 hour per day. Apart from this, the BDDQ also consists of a question asking if the individuals main appearance concern is about weight, with the aim to separate
individuals with eating disorders from those with BDD. Thus, if these requirements are met, and the individual's main appearance concern is not about weight, then the individual may be at risk of having BDD.
The BDDQ also contains a descriptive question where the participant is asked to list the body areas he or she doesn't like. This question was revised by the authors of the current study together with the bigger research project, to better fit the purpose of the current study. In the original form of the BDDQ, the sub-question where the participants are asked to list their body areas of concern is an open-ended question, with associated examples of body parts commonly concerned about. The open-ended question entails that the participants writes the answer in his or her own words. That open-ended sub-question was remade into a
she doesn't like. The alternatives of body areas that the participant could choose where made from the associated examples that were connected to the open-ended question. Example of body areas listed in the BDDQ was the skin, hair and shape or size of the nose (Phillips, 2005). The body areas included in the multiple-choice question, were also the same areas that were used for the BDDQ by a previous study (Petersson & Tuupanen, 2020). Thus, the multiple-choice question aimed to collect quantitative data instead of qualitative data.
Some questions in the BDDQ were not included in the current survey. The questions that weren’t included were open-ended questions connected to previous answers the
individual had given in the scale. If he or she answered yes to some of the questions in the BDDQ, they were asked to specify (e.g., how the appearance concerns had affected them). Theopen-ended questions were excluded because they were not part of the requirements to screen for BDD in the BDDQ. In addition, they were also excluded because the aim of the study was to collect quantitative data and not qualitative data. In a previous evaluation of the BDDQ in a Swedish sample, the same open-ended questions were excluded. The results indicated that the BDDQ was a valid screening tool for BDD (Brohede et al., 2013).
Therefore, only the questions that could be answered with yes or no were included from the BDDQ in the survey.
The BDDQ was shown to have high internal consistency in the current study. The internal consistency was analyzed with KR-20 which resulted in .95. This level was obtained when the descriptive question regarding body areas of concern was excluded. The average score on BDDQ was no and the most frequently reported value was also no. Hence, the BDDQ showed a good internal consistency in the current study. The BDDQ was not affected by any order effects on the participants results based on the different presentations of the scales F(2, 205) = .92, p > .05. See Appendix A for further values for each group, based on the presentation of the scales.
Appearance-Related Repetitive Behaviors Online
The Appearance-Related Repetitive Behaviors Online (ARBO) (Petersson &
Tuupanen, 2020), was used to measure compulsive behaviors typical of BDD but in an online environment. ARBO is a newly developed scale by Petersson & Tuupanen (2020), created to evaluate the presence of checking; avoidance; and reassurance seeking behaviors in an online setting. Also, the scale includes items that screens for manipulation of images. Furthermore, ARBO screens for the severity of these compulsive behaviors online and is divided into two parts with a total of 20 items in the scale.c
The first part of ARBO is the symptom scale including 15 items. The symptom scale attempts to capture compulsive checking behaviors online, that are equivalent to compulsive behaviors that people with BDD do offline (Petersson & Tuupanen, 2020). Furthermore, checking behaviors online are for example represented by items regarding an excessive use of filters and editing tools on photos before they’re uploaded to social media accounts. Examples of questions in ARBO are “How often do you edit pictures of yourself to improve your
appearance before you upload them to social media, because of distress related to your appearance concerns?” and “When you look at pictures of other people online and on social media, how often do you compare parts of your appearance with them?”. The questions in the symptom scale are then answered on a five-point Likert scale, generating total scores that range from 0 to 75 by answering “Never, Seldom, Sometimes, Often or Very often”.
The second part of ARBO is the severity scale, which measures the overall severity of the measured symptoms from the symptom scale. The severity scale consists of two items, screening for related distress caused by the repetitive checking behaviors and distress in connection to the prevention of said behaviors (Petersson & Tuupanen, 2020). Also, apart from the severity questions, another item is included in ARBO that asks the participant to estimate the time spent on engaging in these behaviors online.
Additionally, two other items are included in ARBO to control for the presence of social media use and to screen for the number of pictures posted of the person online. Screening for the presence of social media use, is included to be able to rule out that low scores on questions about social media use are not low due to a lack of social media use (Petersson & Tuupanen, 2020). Whereas the question regarding the number of pictures posted of a person online is asked to assess if people with BDD engage in this behavior. Thus, the appearance-related repetitive behaviors online scale includes a total of 20 items to evaluate checking behaviors in an online setting and their severity. See Appendix B for further presentation of the items in the scale.
ARBO has shown to be a promising scale to measure compulsive behavior online. The psychometric testing of ARBO that has been done so far, has shown that the scale has a solid internal consistency (Petersson & Tuupanen, 2020). Moreover, the scores on the scale are highly correlated with scores on the BDD-SS, which also measures compulsive behaviors found in people with BDD, but in an offline setting. Thus, indicating a good convergent validity of ARBO. Also, this shows promise for ARBO since both scales measure compulsive behaviors but in different settings. Concluding that even though ARBO needs further
psychometric evaluation, it seems to be a promising scale to screen for compulsive behaviors in an online setting.
After the authors of the current study reviewed the existing items in ARBO, an
additional item was added. One item in ARBO is asking participants about a type of web page which does not seem to be popular in Sweden at the current time. The item is: “How often do you use or visit web pages that purports to rate a person's attractiveness based on their
appearance, to get feedback on your appearance concerns?” That item was kept in the scale, but the authors also added the following item: “How often do you visit/use apps or web pages that use likes and comments, to get feedback on your appearance concerns?”. The decision to
add another item was made in an attempt to reflect current popular choices of internet use in Sweden such as Instagram or Facebook, to better capture checking behaviors online. During 2019 and 2020, Instagram and Facebook were the most used social medias in Sweden (Internetstiftelsen, 2020). Hence, a new question was added to ARBO yielding a total of 16 symptom scale questions and a possible total score of 80 in the current study to better represent repetitive online behavior in Sweden.
ARBO was shown to have good reliability in current study. The removal of item number three of the symptom scale in ARBO: “How often do you post pictures of yourself on social media?” resulted in a higher Cronbach’s alpha than when included in the scale.
Subsequently, item number three was also excluded by the creators of ARBO, in the original study for which it was created due to the Cronbach's alpha level being lower when the item was included in the scale (Petersson & Tuupanen, 2020). Moreover, the inclusion of the item that was added by the authors of the current study in an attempt to better reflect current
popular choices of social media sites or apps in Sweden, item 16: “How often do you visit/use apps or web pages that use likes and comments, to get feedback on your appearance
concerns?” resulted in a higher Cronbach’s alpha of the scale, than if it had been excluded. Therefore, the authors decided to keep the additional item in the scale for further analyses. Thus, after these adjustments ABRO symptom scale showed to be highly reliable (α = .89), with scores ranging from 15 to 75 (M = 30.05, SD = 11.09). In addition, the severity scale of ARBO also showed to be highly reliable in the current study as well (α = .82) with scores ranging from 2 up to 20 (M = 5.78, SD = 4.50). Thus, concluding that ARBO was deemed to be a reliable measure in the current study.
The symptom scale and the severity scale of ARBO were also tested for order effects. The symptom scale was not affected by any apparent order effects, F(2, 203) = .50, p > .05. Whereas, the severity scale did show differences between the groups based on the
presentation of the scales (F(2, 205) = .81, p < .05), but a closer inspection of the Games-Howell post-hoc test revealed that no order effect was present. Concluding, that the results on the scales in ARBO were not affected by the order of presentation of the scales. See
Appendix A for further values for each group, based on the presentation of the scales. Body Dysmorphic Disorder Symptom Scale
Body Dysmorphic Disorder Symptom Scale (BDD-SS) was used to measure the presence of risk behavior of BDD and the severity of those behaviors. BDD-SS is a valid and reliable scale with strong internal consistency, appropriate for research at samples like the current study (Wilhelm et al., 2016). Therefore, the BDD-SS was used in the survey for the current study.
The BDD-SS is divided into two parts; one part about the risk behavior of BDD and another part about the severity of the behavior (Wilhelm et al., 2016). The part about
behaviors of BDD consists of 55 items which measure the presence of symptoms, by asking the participants to answer yes or no if they had experience with the presented BDD symptom. The 55 items are divided into seven categories: checking behaviors, grooming behaviors, shape/weight-related behaviors, hair pulling/skin picking behaviors, surgery/dermatology seeking behaviors, avoidance behaviors and BDD-related cognitions. The total points of symptoms can range from 0-55. After each category with symptoms, the question about severity was present. The participants were asked to rate the severity of the symptoms they had experienced on a 0–10-point scale (0 = no problem, 10 = very severe). The total severity of BDD behaviors can range from 0-70. Hence, BDD-SS screens for both risk behavior of BDD and severity of the behavior.
The BDD-SS showed a good reliability in the current study. The BDD-SS symptom scale showed a high reliability (α = .93) with scores ranging from 0 to 46 (M = 16.56, SD = 10.63). Additionally, the severity scale of BDD-SS also showed a good reliability (α = .86)
and scores that ranged from 0 to 70 (M = 15.70, SD = 13.81). Hence, the BDD-SS was deemed a reliable scale for the present study.
Both the symptom scale and the severity scale in BDD-SS were tested for order effects based on the three orders of presentations of the scales. The One-way ANOVA that was run on the symptom scale of BDD-SS showed no differences in scores between the groups based on the order of presentation of the scales, F(2, 189) = 2.51, p > .05. To the contrary, the severity scale did show significant differences between the scores on the scale due to an order effect, F(2, 205) = 4.61, p < .05. The Games Howell post-hoc test revealed that level of distress differed significantly between the second order of the scales (M = 22.1,
SD = 15.05) and third order of presentation of the scales (M = 15.73, SD = 13.83). Meaning
that the participants from the second presentation of the scales showed to experience significantly higher levels of distress due to engagement in risk behavior of BDD, than the participants from the third presentation of the scales. There were no significant differences between the first and the second order of presentation of the scales. See Appendix A for further values for each group, based on the presentation of the scales.
Ethical considerations
The participants took part of the survey freely through the uploaded link online that was posted along with the advertisement for the study. In the advertisement and the pretext of the web survey, it was explained that the collected data were to be included in the current study, a master thesis and in the bigger international research project which aimed to learn more about online behavior, body image and mental health. Participants gave consent to participate in the study by submitting the survey and that information was stated through the pretext in the survey. The pretext also included information that stated that the individual could end all participation at any point, by leaving the web page without submitting the survey. In addition, the participants were recommended to contact their nearest health care
center if any content in the survey had triggered any kind of distress after submitting the survey. Thus, all participants included in the study participated voluntarily and gave their consent by submitting the survey.
The survey was constructed in a program called Artologik to ensure participants confidentiality and anonymity. Artologik is safety certified and ensures a high degree of security in regard to construction of the survey and storage of collected information (Artisan Global Media, 2020). This will ensure that the data collected from participants will remain confidential. Also, to protect participants anonymity and integrity the questions did not include any identifiable features. Hence, the survey was constructed in a program that made sure that the collected data was kept safe and remained confidential.
Analysis
The aim of the present study was to explore if age moderates the relationship between compulsive behavior online and risk behavior of BDD. Before the analyzes were made, some participants were excluded from the study by the authors, because they had only answered a small number of questions in the survey which generated a lot of missing values. First, several analyses were run to evaluate if any participants were at risk of having BDD in the current sample, and to screen for potential differences in behaviors between the participants that were considered to be in high risk of BDD and in low risk of BDD. To determine the estimated amount of participants that were in high risk of BDD, the authors performed a one-sample binomial test on the results from the BDDQ. Next, Pearson's chi-square analysis and Fisher's exact test were used to evaluate demographic differences in the group with high risk for BDD compared to the group with low risk of BDD. Further, Independent sample t-tests were used to compare the scores between the high-risk of BDD group and the low-risk of BDD group on the symptom and severity scales for ARBO and BDD-SS.
Moreover, the authors ran a correlation analysis to explore if compulsive behaviors online and age were associated, and if risk behaviors of BDD and age were associated. The correlation analysis was also done to further establish the association between compulsive behavior online and risk behavior of BDD. The correlation analysis was done by analyzing the sample scores on the symptom scales of ARBO and BDD-SS with age. In addition, we used one-way ANOVAs to compare different age groups on how much they engaged in compulsive online behaviors and risk behavior of BDD. Lastly, a moderation analysis was run to test the hypothesis, if age moderated the relationship between compulsive online behavior and risk behavior of BDD. That is, to see if compulsive behavior online significantly explains parts of the differences in the risk behavior of BDD, and if the association differs with age.
Results
Estimated Risk of Body Dysmorphic Disorder and Appearance Concerns
The estimated risk of BDD within the current study sample was measured by the BDDQ. Figure 1 shows the screening process of the BDDQ. The results showed that more than half of the participants in the current sample (52.9%) expressed concerns regarding their appearance and were therefore qualified for further screening in the BDDQ for risk of BDD. Close to half (46.1%) of the original study sample also reported that they wished that they could think less about their appearance concerns. Although, almost one third of the total participants (31.6%) were then filtered out from the screening process due to the fact that their main appearance concern was related to not being thin enough or afraid of becoming fat. This resulted in a remaining 14.6% that were still thought to be in risk of having BDD.
Furthermore, the participants were asked if certain areas of their lives had been affected negatively due to their appearance concerns, through four sub-questions. To still be included
Figure 1.
Presentation of Responses on the BDDQ
Has it often upset
you a lot?
Has it caused you any
problems with school, work, or other activities?
Has it often gotten in the way of doing things with friends,
dating, your relationships with people, or your social
activities?
Are there things you avoid because of how you look? Yes: 12.1% n = 25 Yes: 6.3% n = 13 Yes: 7.8% n = 16 Yes: 9.22% n = 19
On an average day, how much time do you usually spend thinking about how you look? (Add up all the time you spend in total in a day).
1 - 3 hours: 6%, n = 12 More than 3 hours: 3%, n = 6
At least one hour: Probable BDD 8.7%. n = 18 Note. n = 206
Are you worried about how you look?
Do you think about your appearance problems a lot and wish you could think about them less?
Is your main concern with how you look that you’re not thin enough or that you might get too fat?
No: 47.1% n = 97 Yes: 52.9% n = 109 Yes: 46.1% n = 95 No: 6.8% n = 14 Yes: 31.6% n = 65 No: 14.6% n = 30 No: 1.5% n = 3 Yes: 13.1% n = 27 Less than one hour: 4% n = 9
in the screening for risk of BDD, at least one out of the four areas mentioned in the sub-questions had to have been affected negatively. Among the four sub-sub-questions, the most commonly reported affected area was being upset due to appearance concerns. Resulting in 13.1% of the current sample still being screened for risk of BDD. Finally, the BDDQ asks the participants about how much time they spend on these preoccupations regarding appearance concerns which resulted in 8.7% (n = 18) reporting that they spend more than one hour a day on appearance concerns. Thus, concluding the amount of estimated high-risk BDD cases to be almost a tenth of the current study sample.
To evaluate the results of the BDDQ and attain the number of participants in high risk of BDD, a one-sample binomial test was conducted. A total of 206 participants were included in the screening for risk of BDD by the BDDQ and 18 participants fulfilled the criteria for high risk of BDD. The rates of high risk of BDD for different genders were 6.3% for females
(n = 13), and 2.4% for males (n = 5). Resulting in an estimated rate of high risk of BDD in the current sample to be 8,7% (95% confidence interval: 5.3-13.5%). The
participants that screened positive for BDD on the BDDQ were all between 16 to 30 years old. Thus, no participants older than 30 years old, were deemed to be in high risk of a BDD diagnosis.
The BDDQ included a multiple-choice question regarding which body areas that the participants were concerned about. The nose was the most common body area of concern among the high-risk BDD group, where almost two thirds reported being concerned about it (61%). Skin, mouth and buttock were all the second most commonly Table 2
Body Areas of Concern Among the High-Risk BDD Group (n = 18) Body area n % Nose 11 61 Skin 10 56 Mouth 10 56 Buttocks 10 56 Hair 9 50 Lips 9 50 Breasts 7 39 Stomach 7 39 Cheeks 7 39 Jaw 7 39 Eyes 6 33 Legs 6 33 Genitals 6 33 Muscles 5 28 Hips 5 28 Feet 4 22 Back 3 17 Forehead 3 17 Ears 3 17 Other 4 22
reported body areas of concern, where each of these three body areas were reported by more than half of the sample in the high-risk BDD group (56%). The back, forehead and ears were only reported by 17%, making them the least common areas of concern. See Table 2 for a summary of the body areas of concern within the group with high risk of BDD.
The authors performed Chi-square analyses to compare the high-risk group and the low-risk group of BDD on demographic characteristics. Significant differences were found between the two groups regarding sick leave due to appearance concerns. The people in the high-risk group reported to currently be or to previously have been on sick leave due to appearance concerns (12%, 35% respectively) more frequently, than people in the low-risk group (3%, 14%). Significant differences were also found between the high-risk group and the low-risk BDD group concerning work status, where the high-risk group reported being on sick leave (6%) or to being unemployed (6%) more often than the low-risk BDD group (2%, 1%). There was also a significant difference between the two groups regarding whether they had sought some kind of treatment due to their appearance concerns. Additionally, it was significantly more common to seek cosmetic surgery among the high-risk BDD group (33%) than in the low-risk group (6%). No significant differences were found between the two groups at age, gender and civil status. Further presentation of the results from the Chi-square analysis are shown in Table 3. The questions about work status and sick leave due to
appearance concerns had some missing values, as illustrated in the table.
Do the High-Risk BDD Group Differ From the Low-Risk BDD Group on Engagement in Compulsive Behavior Online and Risk Behavior of BDD?
Independent sample t-tests were conducted to evaluate if the high-risk BDD group and the low-risk BDD group differed on how much they engaged in compulsive behavior online and, if they differed on the amount of experienced distress due to the behaviors online. The
results of the Independent t-tests revealed that the high-risk BDD group and the low-risk BDD group differed significantly on how much they engaged in compulsive behavior online (t(26.66) = 7.41, p < .000), measured by the two groups' scores on ARBO. The high-risk
Table 3
Sociodemographic Characteristics of Participants in the High-Risk BDD Group and the Low-Risk BDD Group
Demographics High-risk of BDD Low-risk of BDD
n % n % Gender Male 5 28 49 26 Female 13 72 137 73 Other 0 0 2 1 Agea 16–24 9 50 83 44 25–34 9 50 76 40 35+ 0 0 29 15 Civil statusa Married 2 11 13 7 Single 8 44 98 52 Cohabitant 2 11 38 20 In a relationship 6 33 37 20 Other 0 0 1 1 Work status*a Student 15 83 116 62
Full time employment 1 6 56 31 Part time employment 0 0 7 4
On sick leave 1 6 3 2
Unemployed 1 6 1 1
Other 0 0 3 2
Sick leave due to appearance concerns*a
Yes, I am on sick leave 2 12 5 3
Yes, previously 6 35 26 14
No, never 9 53 156 83
Sought treatment due to appearance concerns***a
Yes, surgical cosmetic 6 33 12 6 Yes, psychological 2 11 22 12
Yes, medical 5 28 33 18
No, never 5 28 121 64
Note. N = 206. n = 18 in the high-risk of BDD group and, n = 188 in the low-risk of BDD group. Work status: n = 187 participants in the low-risk of BDD group. Sick leave due to appearance concerns: n = 17 participants in the high-risk of BDD group; n = 187 participants in the low-risk of BDD group.
*p ≤ .05, **p ≤ .01, ***p ≤ .001
BDD group engaged in more compulsive behavior online (M = 41.83, SD = 6.62) than the low-risk BDD group (M = 28.88, SD = 10.77). Furthermore, results of another Independent t-test showed that the two groups also differed in distress, related to the compulsive behavior online. The high-risk BDD group experienced more distress (M = 11.56, SD = 4.29) than the low-risk BDD group (M = 5.23, SD = 4.12), and the difference was significant, t(204) = 6.20,
p < .000. Hence, the high-risk BDD group reported performing more compulsive behavior
online and also experienced more distress due to those behaviors, than the low-risk BDD group.
Another Independent sample t-test was conducted to explore if the group considered to be in high risk of BDD and the group considered to be in low risk of BDD showed
different results on how much they engaged in risk behavior of BDD. Results from the
Independent sample t-tests showed that the high-risk group and the low-risk group differed on the amount of risk behavior of BDD, measured by the BDD-SS. The high-risk BDD group engaged in more risk behavior (M = 29.25, SD = 7.08) than the group with low risk of BDD (M = 15.39, SD = 10.14), and the difference was significant, t(188) = 5,34, p < .000. Also, results from an additional t-test showed that the distress that the two groups experienced due to the performance of the risk behavior of BDD differed as well. The high-risk BDD group experienced more distress (M = 28.89, SD = 7.86) than the low-risk BDD group (M = 14.44,
SD = 13.61). The difference in experienced distress between the two groups was significant, t(27.95) = 6.88, p < .000. Thus, the high-risk of BDD group engaged in more risk behavior of
BDD and also reported that they experienced more distress due to those behavior, than the participants in the low-risk of BDD group.
The Association Between Compulsive Behavior Online, Risk Behavior of BDD and Age The authors conducted a correlation analysis to explore the associations between compulsive online behavior as measured by scores on ARBO, risk behavior of BDD
measured by scores on BDD-SS and age. The results of a Pearson's product-moment
correlation showed a significant negative correlation between age and scores on the symptom scale in ARBO, (r = -.35, p < .000). Meaning that, lower age was associated with more
compulsive online behavior. Furthermore, the results from Pearson's product moment correlation also showed a significant negative association between age and symptom scores on BDD-SS, (r = -.30, p <.000). Meaning that, younger age was also associated with risk
behavior of BDD. Lastly, as expected, the results from the correlation analysis showed a strong, positive correlation between scores on the symptom scales of ARBO and BDD-SS,
r = .78, p < .000. That is, people who engage in more compulsive behaviors online also
engage in more risk behaviors of BDD.
Analyses of Age Groups on Compulsive Behavior Online and Risk Behavior of BDD To uncover if the three age groups (group 16-24 year; group 25-34 years; and group 35+ years) differed in the amount of compulsive behavior online and on the level of risk behavior of BDD, several one-way ANOVAs were run. The results are presented in Table 4. The first one-way ANOVA measured the ARBO Symptom Scale and the age groups. As expected, all the age groups differed significantly on their scores on ARBO. A closer
inspection of the Games-Howell post-hoc test showed that participants aged 16-24 years old engaged in the most compulsive behavior online compared to the other two age groups. In turn, participants aged between 25-34 years old engaged in more compulsive behavior online,
than participants aged 35 and older. Concluding that, as age increased, the online compulsive behavior decreased between the age groups.
A one-way ANOVA was also used to evaluate the experienced distress reported by the participants, due to performance of compulsive online behavior in relation to age. The results showed that there were significant differences between the age groups regarding distress. The group with participants aged 16-24, experienced significantly more distress than the two older age groups. The group with 25–34-year-olds and the group with participants aged 35 and older did not differ significantly. Thus, the participants aged 16-24 experienced more distress due to compulsive behavior online, than participants aged 25 or older.
Next, an additional one-way ANOVA was conducted to uncover if the participants differed in how much they engaged in risk behaviors of BDD, depending on what age group they belonged to. The results showed that the age groups differed significantly on their scores. The Games-Howell post hoc test showed that participants aged 16-24 years old showed significantly higher scores on BDD-SS Symptom Scale than both the participants aged between 25-34 and participants aged 35 and older. Although, there were no significant differences between participants aged between 25-34 and participants aged 35 and older. In summary, the age group with the youngest participants showed higher scores on BDD-SS which also indicated they engaged in more risk behaviors of BDD, than people aged 24 and older.
Lastly, a final one-way ANOVA was run to evaluate if the three age groups differed in the amount of experienced distress due to engagement in risk behavior of BDD. That was analyzed with scores on BDD-SS Severity Scale and the age groups. The results revealed significant differences between the age groups regarding experienced distress and, Games-Howell's post hoc test showed that all groups differed between each other. The age group with the youngest participants aged 16-24 years old, felt more distress than the other two age
groups. Also, the group with participants aged between 25-34 felt more distress than the group with participants aged 35 or older. Thus, the experienced distress due engagement in risk behaviors of BDD showed to decrease as age increased.
The Effect of Age on the Relationship Between Compulsive Behavior Online and Risk Behavior of BDD
This study was conducted to examine the relationship between compulsive behaviors online and risk behaviors of BDD, and to investigate if age, as a continuous variable, acts as a moderating variable in this relationship. It was hypothesized that compulsive behaviors online would positively predict risk behaviors of BDD and that age would enhance the relationship. A simple moderation analysis was carried out to test this hypothesis. The results showed that age significantly moderated the relationship between compulsive behavior online and risk behavior of BDD (b = .01, 95% C.I. [-.00, .03], t = 2.08, p < .05). The simple slope analysis revealed that all levels of the moderating variable age had an impact on the level of risk of BDD. Where, low level of age (b = .63, 95% CI [.45, .73] t = 13.11, p < .001), average level of age (b = .77, 95% CI [.66, .89], t = 13.65, p < .001) and, also high level of age (b = .92, 95% CI [.69, 1.15], t = 7.85, p < .001), significantly enhanced the impact of compulsive online behavior on risk behavior of BDD (where low level of age was 15,89, average level was 26.85, and high level was 37.81). Hence, the results of the analyses suggest that people who engage in more compulsive behavior online are more likely to also engage in higher
Table 5
Moderator analysis: Compulsive Behaviors Online (ARBO), Age and Risk Behaviors of BDD (BDD-SS)
Estimate SE(HC3) 95%CI p
LL UL
Intercept 17.19 .60 16.00 18.38 <.001
ARBO .77 .06 .66 .89 <.001
Age .04 .08 -.12 .20 .603
ARBO x Age .01 .01 .00 .03 .039
Note. Number of studies = 188. CI = confidence interval; LL = lower limit; UL = upper limit. SE(HC3) = heteroscedasticity consistent standard errors
levels of risk behavior of BDD, this was especially true for participants with an older age. Table 5 reveals the results from the moderation analysis and Figure 2 further illustrates the results from the simple slope analysis.
Discussion
The present study was conducted to extend the research on the relationship between compulsive behavior online and risk behavior of BDD, and if that relationship differs depending on age. The hypothesis was that people who engaged in more compulsive
behaviors online also would engage in more risk behavior of BDD and that this relationship would be affected by age.The findings in the present study showed that participants of all ages in the current sample engaged in the investigated behaviors both offline and online. Additionally, it was also found that the behaviors differed between age groups on how much the participants engaged in compulsive behaviors and mental acts online and offline.
24, engaged the most in these behaviors and that the behaviors online declined with age. Regarding the offline context, the group with the youngest participants also performed the most risk behaviors of BDD, compared to the other groups contained of older participants. However, the two groups with older participants did not differ in how much they engaged in risk behaviors of BDD in an offline setting. Furthermore, as hypothesized, age was found to affect the relationship between compulsive online behavior and risk behavior of BDD. In addition, it was found that if older participants engaged in a lot of compulsive online
behaviors, then it predicted more risk behaviors of BDD than what was found for adolescents and young adults that engaged in the same online behaviors.
Discussion of Findings in the Current Study
There is a possible explanation for the finding that older participants who engaged in more compulsive behaviors online also corresponded with the highest level of risk behavior of BDD in the current study. An explanation may be that when older participants engage in these compulsive behaviors online, the behaviors are more indicative of BDD or other body image related issues, than when younger people engage in the same behaviors online. This would be supported by our results showing that the older participants engaged less in compulsive behaviors online and in risk behaviors of BDD in general, than younger
participants. Further, statistics show that adolescents and young adults are the most common users of social media (SCB, 2020). Meaning that older people are less frequent users of social media in general. Thus, if older individuals engage in a lot of compulsive online behavior, then they may be at greater risk of having BDD, than when adolescents and young adults engage in the same behavior.
As expected, the current study also found that more compulsive behavior online corresponded with more risk behavior of BDD, especially in older participants, thus indicating that ARBO could be a good predictor of risk behaviors of BDD. Also, the
behaviors online and offline showed to correlate significantly and, thereby suggesting that they do indeed measure similar things but in different contexts. Where, higher scores on ARBO meant engaging in more compulsive behaviors online, and higher scores on BDD-SS meant engaging in more risk behavior of BDD offline. This would also be explained by previous research regarding social media and body image that show that appearance-focused behaviors online relate to body dissatisfaction (see for example, Seekis et al., 2020). In addition, certain comparative behaviors online have also been shown to be risk factors for generating a bad body image (Salomon & Brown, 2018), which is in line with research building on the social comparison theory (Tiggemann & Anderberg, 2019). Further, because more time is being spent online and especially on social media, it is not surprising that compulsive behaviors typical of BDD appear online as well as offline. Hence, our findings suggest that ARBO was a good predictor of body dysmorphic symptoms (i.e., risk of BDD) in the current study.
In the BDDQ, many participants were filtered out from the screening process for being considered to be at high risk of having BDD, because their appearance concerns were connected to being thin, even though they showed sub-clinical symptoms of BDD. Also, the results of the study showed that more compulsive behaviors online were strongly related to risk behavior of BDD but, only a small portion of the sample fulfilled the criteria of being in high risk of having BDD according to the results of the BDDQ. Furthermore, even though the high-risk BDD group engaged the most in compulsive behaviors online, the group with low risk of BDD engaged in the same behaviors as well. In the BDDQ, the item that ruled out many participants from being considered to be in high risk of having BDD, was about their main concern being related to weight issues. Concerns about food and weight issues are common concerns in people with an eating disorder (APA, 2013). Also, people with an eating disorder have shown to engage more in body checking behaviors than healthy controls and
